Date post: | 14-Mar-2018 |
Category: |
Documents |
Upload: | trinhquynh |
View: | 221 times |
Download: | 2 times |
Molecular Cell, Volume 60
Supplemental Information
Mitochondrial Phosphoenolpyruvate Carboxykinase Regulates Metabolic Adaptation and Enables Glucose-Independent Tumor Growth Emma E. Vincent, Alexey Sergushichev, Takla Griss, Marie-Claude Gingras, Bozena Samborska, Thierry Ntimbane, Paula P. Coelho, Julianna Blagih, Thomas C. Raissi, Luc Choinière, Gaëlle Bridon, Ekaterina Loginicheva, Breanna R. Flynn, Elaine C. Thomas, Jeremy M. Tavaré, Daina Avizonis, Arnim Pause, Douglas J.E. Elder, Maxim N. Artyomov, and Russell G. Jones
Inventory of Supplemental Information:
Figure S1, related to Fig. 1. Metabolite analysis of A549 cells grown in the absence of glucose. Figure S2, related to Fig. 2. Glucose withdrawal redirects glutamine metabolism to maintain the TCA cycle and for the production of PEP. Figure S3, related to Fig. 3. TCA cycle metabolism in the absence of PCK2. Figure S4, related to Fig. 4. Serine, glycine and ATP are made from glutamine upon glucose withdrawal in a PCK2-dependent manner Figure S5, related to Fig. 5. Cancer cells require PCK2 to maintain glucose-independent proliferation and tumor growth in vivo. Figure S6, related to Fig. 6. PCK2 expression is regulated by glucose availability. Figure S7, related to Fig.7. PCK2 expression is elevated in NSCLC. Table S1, List of qPCR primers Supplementary Experimental Procedures
Supplementary References
Figure S1, related to Fig. 1. Metabolite analysis of A549 cells grown in the absence of glucose
Arginine
Asparagine
Aspartate
Pyroglutamate
TryptophaneSerin
eValin
e
Succinate
Phenylalanine
Glutamine
Threonine
Tyrosine
Inosine5-monoP
Leucine
Glycine
Isoleucine
Cysteine
Alanine
Histidine
MethionineAMP
Cytidine5-diP
U5-monoP
GlutamateGMP
PhosphoenolpyruvateADP
GDP
Glutamic acidGTP
NAD
Cytidine5-tri
P
Gluthatione ox
ATPNADP
Betaine
Hypotaurine
Adenylate Energy Charge (AEC)
Cytidine5-m
onoPU5-tri
PProlin
e
Ribulose 5P
CystathionineSAM
U5diP-N-Acetyl Glucosamine
Ribose 5P
TDP pos
TTP pos
U5-diP-Glucuronic acidCitra
te
Creatine
U5-diP
a-KetoglutarateMalate
Pyruvate
Taurine
Creatinine
Fumarate
P-creatine
U5-diP-Glucose
Lactate-5
-4
-3
-2
-1
0
1
2
3
4
5
Arg
inin
e
Asp
arag
ine
A
spar
tate
P
yrog
luta
mat
e
Tryp
toph
ane
S
erin
e
Valin
e
Suc
cina
te
Phe
nyla
lani
ne
Glu
tam
ine
Th
reon
ine
Ty
rosi
ne
Inos
ine5
-mon
oP
Leuc
ine
G
lyci
ne
Isol
euci
ne
Cys
tein
e
Ala
nine
H
istid
ine
M
ethi
onin
e
AM
P
Cyt
idin
e5-d
iP
U5-
mon
oP
Glu
tam
ate
G
MP
P
hosp
hoen
olpy
ruva
te
AD
P
GD
P
Glu
tam
ic a
cid
G
TP
NA
D
Cyt
idin
e5-tr
iP
Glu
that
ione
ox
AT
P
NA
DP
B
etai
ne
Hyp
otau
rine
A
EC
C
ytid
ine5
-mon
oP
U5-
triP
P
rolin
e
Rib
ulos
e 5P
C
ysta
thio
nine
S
AM
U
5diP
-N-A
cety
l Glu
cosa
min
e
Rib
ose
5P
TDP
pos
TT
P po
s
U5-
diP
-Glu
curo
nic
acid
C
itrat
e
Cre
atin
e
U5-
diP
a-
Ket
oglu
tara
te
Mal
ate
P
yruv
ate
Ta
urin
e
Cre
atin
ine
Fu
mar
ate
P
-cre
atin
e
U5-
diP
-Glu
cose
La
ctat
e -5
-4
-3
-2
-1
0
1
2
3
4
5
log 2
(fol
d ch
ange
in m
etab
olite
abu
ndan
ce)
(-G
lc re
lativ
e to
+G
lc)
* *
* * * * * * * * *
* * * * * * * * * * * * * * * *
* *
* * *
* * * * * * * *
* *
Vincent_FigS1
Figure S2, related to Fig. 2. Glucose withdrawal redirects glutamine metabolism to maintain the TCA cycle and for the production of PEP.
010
020
030
040
050
0Citrate NGCitrate FG
Malate NGMalate FG
Fumarate NGFumarate FG
Succinate NGSuccinate FG
aKG NGaKG FG
Glutamate NG Glutamate FG
Vincent_FigS2
Glutamate
Succinate
α-KG
Fumarate
Malate
Relative metabolite abundance (% of +Glc)
+ -
Citrate
+ -
+ -
Aspartate + -
+ -
+ -
+ -
0
A
B
PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend
13C 12C
100 300 200 400 500
0 1000 2000 3000Aspartate NGAspartate FG
0 1000 3000 2000
FG 6H
NG 6H
0
2
4
6
8
10
Pyr
PEP
Lac
0
20
40
60
80
100
C
0
10
-Glc +Glc
Rel
. Abu
ndan
ce
U-[13C]-Q Lactate
PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend
13C 12C
8
6
2
4
0
100
80
60
20
40
% o
f poo
l (in
-Glc
)
PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegendm+0 m+3
Pyruvate PEP Lactate
D U-[13C]-Q Lactate
Figure S3, related to Fig. 3. TCA cycle metabolism in the absence of PCK2.
0 10 20 40 120
360
0
20
40
60
Vincent_FigS3
D
FG EV
NG EV
FG 6
NG 6
FG 8
NG 8
0
5
10
1530
20
10
0 Glc: + - + - + -
shRNA: Con PCK2(1) PCK2(2)
13C
-citr
ate
m+6
(% o
f poo
l) U-[13C]-Q Citrate (m+6)
A549
C
010
020
030
0
Citrate NG 8Citrate FG 8
Citrate NGCitrate FG
Malate NG 8Malate FG 8
Malate NGMalate FG
Fumarate NG 8Fumarate FG 8
Fumarate NGFumarate FG
Succinate NG 8Succinate FG 8
Succinate NGSuccinate FG
aKG NG 8aKG FG 8
aKG NGaKG FG
Glutamate NG 8Glutamate FG 8Glutamate NG Glutamate FG
Glutamate
Succinate
α-KG
Fumarate
Malate
Aspartate
Relative metabolite abundance (% of +Glc)
Citrate
100
+ - + -
Con PCK2
+ - + -
Con PCK2
+ - + -
Con PCK2
+ - + -
Con PCK2
+ - + -
Con PCK2
+ - + -
Con PCK2
+ - + -
Con PCK2
PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend
13C 12C
200 300 0
0 200 400 600Aspartate NG 8Aspartate FG 8
Aspartate NGAspartate FG
200 400 600 0
30
20
10
0
40
13C
-citr
ate
m+6
(% o
f poo
l)
FG EV
NG EV
FG 6
NG 6
FG 8
NG 8
0
10
20
30
40
Glc: + - + - + - shRNA: Con PCK2(1) PCK2(2)
H1299
0 10 20 40 120 360 Time (min)
0
40
60
13C
-PE
P (%
of p
ool)
20
0102040120
360
0
2
4
6
8
Ev68
Con shRNA:
PCK2(1) PCK2(2)
A B
PEPCK (total) Actin
Glc: - + - + - + siRNA: PCK1 PCK2 Con
Figure S4, related to Fig. 4. Serine, glycine and ATP are made from glutamine upon glucose withdrawal in a PCK2-dependent manner.
FG 12h
6hP
NG 12h
6P
0
1
2
3
4
FG Con
NG Con
FG ENO
NG ENO
FG PH
NG PH0.0000
0.0001
0.0002
0.0003
0.0004
0.0005
Vincent_FigS4
H G 5
Glc: + - siRNA: PHGDH
Rel
. Abu
ndan
ce (x
10-2
)
4
3
2
1
0 + -
ENO1 + - Control
FG NG0.0
0.5
1.0
1.5
2.0
2.5
A B
FG NG0
2
4
6
8
101.5
0
1.0
-Glc +Glc
Rel
. Abu
ndan
ce
U-[13C]-Q Serine PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend
13C 12C
U-[13C]-Q Glycine
-Glc +Glc
0.5
3
0
2
4 R
el. A
bund
ance
1
2.0
0
1.5
2.5
-Glc +Glc
Rel
. Abu
ndan
ce
U-[13C]-Q Serine PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend
13C 12C
U-[13C]-Q Glycine
-Glc +Glc
1.0
8
0
6
10
Rel
. Abu
ndan
ce
2 0.5 4
U-[13C]-Q Serine
A549 H1299
α-Enolase
siRNA:
Actin
PHGDH
siRNA:
Actin
C D
E F
1 20.0
0.5
1.0
1.5
m+0
m+1
m+2
m+3
0
20
40
60
80
100
m+0
m+1
m+2
0
20
40
60
80
100
0 20
100
% o
f ser
ine
pool
PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend+Glc -Glc
60 80
40
0 20
100
60 80
40
% o
f gly
cine
poo
l
A549 A549 PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend+Glc -Glc
0 10 20 40 120
360
0
2
4
6
8
Ev68
0 10 20 40 120
360
0.0
0.5
1.0
1.5
2.0
2.5
Ev68
0 10 20 40 120 360 Time (min)
0
2.0 2.5
13C
-Gly
cine
(% o
f poo
l)
1.0 1.5
0 10 20 40 120 360 Time (min)
0
6
8
13C
-Ser
ine
(% o
f poo
l)
2
4
0.5
0102040120
360
0
2
4
6
8
Ev68
Con shRNA:
PCK2(1) PCK2(2)
0102040120
360
0
2
4
6
8
Ev68
Con shRNA:
PCK2(1) PCK2(2)
Figure S5, related to Fig. 5. Cancer cells require PCK2 to maintain glucose-independent proliferation and tumor growth in vivo.
Vincent_FigS5
A
Con
Pck2
ENO
PHGDH
0
20
40
60
80
100
120120 100
80 60
0
40 20
siRNA:
PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend+Glc -Glc
% v
iabl
e ce
lls
B
A549 PCK2
Actin
siRNA:
H1299
PCK2
Actin
H1299
E
0 24 48 72 960
500
1000
1500
2000
EV68
0 24 48 72 96 Time (h)
2.0 1.5 1.0 0.5
0
A549
0 24 48 72 960
500
1000
1500
2000
Eno 0mMPHGDH 0mMPck2 0mMCon 0mMPCK2(1)
Control
0 24 48 72 960
500
1000
1500
2000
Eno 0mMPHGDH 0mMPck2 0mMCon 0mM
shRNA:
PCK2(2) C
ell n
umbe
r (x1
03)
PCK2
Actin
Control shRNA: PCK2
C
D
G F Cells injected: Tumors:
PCK2
Actin
Control shRNA: PCK2
Tumors:
shRNA:
PCK2
Actin
A549
Cells injected:
shRNA:
Figure S6, related to Fig. 6. PCK2 expression is regulated by glucose availability.
Vincent_FigS6
HIF1A siRNA EPAS1 siRNA
-
-
-
-
- -
- -
+ +
+ +
+
+
+
+
1.0
0
Rel
. mR
NA
expr
essi
on
(rel
ativ
e to
con
trol)
0.25
0.50
0.75
1.5
EPAS1
*** ***
*** ***
B PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend+Glc -Glc
1.0
0
0.50
HIF1A siRNA: EPAS1 siRNA:
- -
- -
+ + + +
PCK2
Fold
exp
ress
ion
(rel
ativ
e to
con
trol) **
0.25
0.75
C
HIF1A siRNA EPAS1 siRNA
-
-
-
-
- -
- -
+ +
+ +
+
+
+
+
1.0
0
Rel
. mR
NA
expr
essi
on
(rel
ativ
e to
con
trol)
0.25
0.50
0.75
1.5
HIF1A
*** *** *** ***
PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegend+Glc -Glc
A549
SUCLA2
Actin
siRNA:
SUCLG2
Actin
A
Figure S7, related to Fig.7. PCK2 expression is increased in NSCLC.
Vincent_FigS7
24 35 17 32 15 29 9 33 3 19 6 30 31 22 5 4 13 7 11 14 36 37 34 28 27 25 21 10 160
5
10
15
20
25
30
5
10
15
20
25
30
0
PC
K2
prot
ein
expr
essi
on (A
U)
Patient number
* * * * *
*
*
* *
*
*
*
*
*
*
*
*
*
PEP all
0 10 20 40 120
360
0.000
0.002
0.004
0.006
0.008
0.010LegendLegendNormal Tumor
Supplementary Figure Legends:
Figure S1, related to Fig. 1. Metabolite analysis of A549 cells grown in the absence of
glucose.
A549 cells were cultured in the presence (+Glc) or absence of glucose (-Glc) for 48h and
metabolite abundances determined using both LC- and GC-MS. The data are displayed as the
log2 of the fold change in metabolite abundance between –Glc relative to +Glc conditions. Data
are represented as mean +/- SEM of three independent cultures.
Figure S2, related to Fig. 2. Glucose withdrawal redirects glutamine metabolism to
maintain the TCA cycle and for the production of PEP.
(A) Relative abundances of glutamate, α-KG, succinate, fumarate, malate, citrate and aspartate
after culture of A549 cells for 12h in the presence or absence of unlabeled glucose (Glc). Cells
were cultured with U-[13C]-Q for the last 6h of the 12h incubation. Data are presented relative to
metabolite abundance in glucose-replete conditions (+Glc).
(B) Model for U-[13C]-Q flux in the TCA cycle under glucose withdrawal. In the absence of
glucose fully labeled TCA cycle intermediates made from U-[13C]-Q can be used to make fully
labeled pyruvate and PEP (m+3). Pyruvate m+3 can re-enter the TCA cycle, as acetyl-CoA
(m+2), which condenses with fully labeled OAA (m+4) to give citrate m+6. Abbreviation: PEP,
phosphoenolpyruvate.
(C) Relative abundance of lactate in A549 cells cultured with U-[13C]-Q as in (A).
(D) The contribution of mass isotopologues (unlabeled m+0 or fully-labeled m+3) to the total
pyruvate, PEP and lactate pools in A549 cells cultured with U-[13C]-Q as in (A).
Data are represented as mean +/- SEM of three independent cultures.
Figure S3, related to Fig. 3. TCA cycle metabolism in the absence of PCK2.
(A) Immunoblot for total PEPCK and actin on lysates from A549 cells transfected with either
control siRNA or siRNA targeting either PCK1 or PCK2. Cells were incubated in the presence or
absence of glucose for 48h.
(B) Contribution of U-[13C]-Q to PEP in A549 cells expressing either control shRNA or shRNA
targeting PCK2 in glucose free conditions. Cells were pre-incubated in the absence of glucose for
6h before incubation with U-[13C]-Q (time 0). Cells extracts were harvested at the time points
shown.
(C) Relative abundances of glutamate, α-KG, succinate, fumarate, malate, citrate and aspartate in
A549 cells expressing control or PCK2 shRNA. Cells were cultured in the presence or absence of
unlabeled glucose for 12h, and cultured with U-[13C]-Q for the last 6h of the 12h incubation. Data
are made relative to metabolite abundance in glucose-replete conditions (+Glc).
(D) Proportion of mass isotopologue citrate m+6 in A549 and H1299 cells expressing control or
PCK2 shRNA. Cells were cultured in the presence or absence of glucose as in (C).
(B-D) Data are represented as mean +/- SEM of three independent cultures.
Figure S4, related to Fig. 4. Serine, glycine and ATP are made from glutamine upon glucose
withdrawal in a PCK2-dependent manner.
(A-B) Relative abundances of serine and glycine and the contribution of U-[13C]-Q to these
metabolites in A549 cells (A) or H1299 cells (B). Cells were cultured in the absence or presence
of unlabeled glucose for 12h, with U-[13C]-Q being present for the last 6h of the 12h incubation.
(C-D) Contribution of the mass isotopologues of serine (C) and glycine (D) to the total
metabolite pool in A549 cells. Cells were incubated with U-[13C]-Q in the presence or absence of
glucose for 48h before cells extracts were harvested.
(E-F) Contribution of U-[13C]-Q to serine (E) and glycine (F) in A549 cells expressing either
control shRNA or shRNA targeting PCK2 in glucose free conditions. Cells were pre-incubated in
the absence of glucose for 6h before incubation with U-[13C]-Q (time 0). Cells extracts were
harvested at the time points shown.
(G) Immunoblot for enolase, PHGDH and actin on lysates from A549 cells 3 days post
transfection with either control siRNA or siRNA targeting either ENO1 or PHGDH.
(H) Relative abundance of 13C-serine in A549 cells transfected with either control siRNA or
siRNA targeting either ENO1 or PHGDH. Cells were cultured as in (A).
(A-F and H) Data are represented as mean +/- SEM of three independent cultures
Figure S5, related to Fig. 5. Cancer cells require PCK2 to maintain glucose-independent
proliferation and tumor growth in vivo.
(A) Immunoblot for PCK2 and actin on lysates from A549 and H1299 cells 3 days post
transfection with either control siRNA or siRNA targeting PCK2.
(B) Proliferation of A549 cells expressing either control shRNA or shRNA targeting PCK2 in
glucose-free media over 96 hours. Data are represented as mean +/- SEM for biological replicates
(n=5).
(C) A549 cells transfected with siRNA targeting PCK2, ENO1 or PHGDH were cultured for 48h
in the presence or absence of glucose before staining with propidium iodide to measure the
percentage of viable cells. Data are represented as mean +/- SEM of three independent cultures.
(D-G) Immunoblot for PCK2 and actin on lysates from A549 (D) and H1299 (F) cells, expressing
either control or shRNA targeting PCK2, injected into the flanks of nude mice. Immunoblot for
PCK2 and actin on lysates from the resulting tumors (where enough tissue was available) formed
from the A549 (E) and H1299 (G) cells injected.
Figure S6, related to Fig. 6. PCK2 expression is regulated by glucose availability.
(A) Immunoblot for SUCLG2, SUCLA2 and actin on lysates from A549 cells 3 days post
transfection with either control siRNA or siRNA targeting either SUCLG2 or SUCLA2.
(B) Expression of HIF1A and EPAS1 mRNA in A549 cells. A549 cells transfected with control
siRNA or siRNA targeting HIF1A, EPAS1 or a combination of both were incubated in the
presence or absence of glucose for 48h. Transcript levels for HIF1A or EPAS1 mRNA were
determined relative to OGDH mRNA levels, and normalized relative to control cells (transfected
with control siRNA). ***, p<0.001.
(C) Relative expression of PCK2 mRNA as determined by qPCR. A549 cells transfected with
siRNA against HIF1A, EPAS1 or a combination of both were cultured in the presence or absence
of glucose for 48h. PCK2 mRNA transcript levels were determined relative to OGDH mRNA
levels, and normalized relative to expression in cells transfected with control siRNA. **, p<0.01.
(B-C) Data are represented as mean +/- SEM of three independent cultures.
Figure S7, related to Fig.7. PCK2 expression is increased in NSCLC.
Quantified proteomic data from 29 NSCLC patients. Each bar represents the average expression
of PCK2 in 3 pieces of normal tissue (black bars) or 3 pieces of tumour tissue (white bars) for
each patient (mean +/- SEM). The strength of evidence for a difference in expression between the
normal and tumour samples was determined by a Kruskal-Wallis test, *, p<0.05.
Table S1: List of qPCR primers
Gene Forward Primer Reverse Primer PCK2 TGCCAGGCTGGAAAGTGGAGTGT GCAACCCCAAAGAAGCCGTTCTCA HIF1A TGCTCATCAGTTGCCACTTCC CGCTGTGTGTTTTGTTCTTTACCC EPAS1 GCCTCCATCATGCGACTGGCA CCATCTTGGGTCACCACGGCA OGDH AGACCCCTGGGATCATGCAGTTCA CCTCGCAGCCTTCTAGACCAAAGC PCK1 CTGTGACGGCTCTGAGGAGGAGA
A CCACATCCCTGGGGTCAGTGAGAG
Supplementary Experimental Procedures:
Materials
The PEPCK inhibitor 3-mercaptopicolinic acid (MPA) was obtained from Santa Cruz
Biotechnology (Dallas, TX, USA). Primary antibodies to PCK2 and actin, as well as HRP-
conjugated anti-rabbit and anti-mouse secondary antibodies were obtained from Cell Signaling
Technology (Danvers, MA, USA). Primary antibody to α-enolase was obtained from Santa Cruz
Biotechnology (Dallas, TX, USA). PHGDH antibodies were obtained from Thermo Fisher
(Waltham, MA, USA). Primary antibodies to SUCLG2 and SUCLA2 were obtained from Novus
Biologicals (Littleton, CO, USA).
Cell culture
A549 and H1299 cell lines were obtained from ATCC (Manassas, VA, USA). Cells were
cultured in ‘growth medium’ (DMEM (A549 cells) or RPMI (H1299 cells)) supplemented with
10% fetal bovine serum (FBS), 20000U/ml penicillin, 7mM streptomycin and 2mM glutamine
and non-essential amino acids (for H1299 cells). Cells were grown at 37°C in a humidified
atmosphere supplemented with 5% (v/v) CO2. For experiments cells were cultured in DMEM
with 20000U/ml penicillin, 7mM streptomycin and 10% dialysed FBS (Wisent, Saint-John-
Baptiste, QC, Canada), 2mM glutamine and 25mM glucose were then added as required.
shRNA and siRNA knockdown
PCK2 knockdown was achieved using lentiviral shRNA vectors (ID numbers: TRCN0000052664
and TRCN0000052666) from the TRC shRNA collection (Sigma-Aldrich, St. Louis, MO).
Lentiviral supernatants were generated as described (Huang et al., 2012). All transient
knockdowns in this manuscript were achieved using the SMARTpool ON-TARGETplus siRNA
reagent (composed of 4 individual siRNAs for each target) from GE Dharmacon (Layette, CO,
USA). 50nM siRNA was incubated for 20min with RNAimax in OptiMEM medium (Life
Technologies) in a 12 well plate, followed by seeding of cells (60,000 cells/well). Cells were
transfected a second time 48 hours later, and cells plated for assays the following day.
RNA-Seq analysis
A549 cells were cultured in the presence (25 mM) or absence of glucose for 48 hours prior to
RNA extraction. For cDNA synthesis, custom oligo-dT primers were used with a barcode and
adapter-linker sequence (CCTACACGACGCTCTTCCGATCT—XXXXXXXX-T15). After first
strand synthesis, samples were pooled together based on ACTB qPCR values, and RNA-DNA
hybrids degraded using consecutive acid-alkali treatment. A second sequencing linker
(AGATCGGAAGAGCACACGTCTG) was ligated using T4 ligase (NEB) followed by SPRI
clean-up. The mixture was then PCR enriched for 12 cycles and SPRI purified to yield final
strand specific RNA-seq libraries as previously described (Jha et al., 2015). Libraries were
sequenced using a HiSeq 2500 (Illumina) using 50bpX25bp pair-end sequencing. Second mate
was used for sample demultiplexing, at which point individual single-end fastqs were aligned to
mm9 genome using TopHat. and gene expression was obtained using ht-seq and DESeq2 for
differential expression. Raw and processed sequencing data are deposited at Pubmed GEO under
GSE66556.
Metabolite profiling by LC-MS
Liquid chromatography was performed using a 1290 Infinity ultra-performance LC system
(Agilent Technologies, Santa Clara, CA, USA) equipped with a Scherzo 3 μm, 3.0×150mm SM-
C18 column (Imtakt Corp, Japan). The column temperature was maintained at 10°C and the
mobile phases A and B consisted of water containing 5 and 200mM ammonium acetate,
respectively. The chromatographic gradient started at 100% mobile phase (A) with a 5 min
gradient to 100% (B). This was followed by a 5min hold time at 100% mobile phase B at a flow
rate of 0.4ml/min. A subsequent re-equilibration time (6min) was performed before the next
injection. Sample volumes of 5μl were injected for LC-MS analysis. LC-MS analysis was
performed on an Agilent 6540 UHD Accurate-Mass Q-TOF mass spectrometer (Agilent
Technologies, Santa Clara, CA, USA). Analyte ionization was accomplished using an
electrospray ionization source (ESI) in both positive and negative polarities. The source operating
conditions were set at 325°C and 9l/min for gas temperature and flow respectively, nebulizer
pressure was set at 40psi and capillary voltage was set a 4.0kV. Reference masses 121.0509,
922.0099, 1033.9881 were introduced into the source through a secondary spray nozzle to ensure
accurate mass. MS data were acquired in full scan mode mass range: m/z 100-1000; scan time:
1.4s; data collection: centroid and profile. Retention times, and accurate mass for each compound
were confirmed against authentic standards as well as matched unlabeled cell extracts grown
under the same conditions when cells were undergoing stable isotope tracer analysis (SITA).
Data were quantified by integrating the area underneath the curve of each compound using
MassHunter Qual (Agilent Technologies, Santa Clara, CA, USA). Each metabolite’s accurate
mass ion and subsequent isotopic ions were extracted (EIC) using a 10 ppm window.
Network-based data integration
Network-based integration of metabolite and gene expression datasets was conducted as
previously described (Jha et al., 2015). A chemical mapping table between carbon atoms in
substrates and products for all annotated reactions in the KEGG database using RPAIRs entries
was made (Kanehisa et al., 2012). Only those atoms that could be uniquely identified by the
chemical structure (i.e. constant between different stereoisomers) were used. Using such a
chemical mapping table we built a network of individual carbon atoms connected by a substrate-
product relationship via biochemical reactions. Each metabolite is present in the network through
a number of disconnected nodes representing individual carbon atoms within the metabolite, with
common metabolites (i.e. water, ATP, NAD) removed from the network to avoid redundancy. At
the final filtering stage only reactions catalyzed by enzymes with significant expression levels
were kept. The top 12000 expressed genes were used as a cutoff, to generate a network
containing approximately 6000 nodes and 6000 edges.
To search for sub-networks upregulated by glucose deprivation, network nodes and edges
were assigned a score by differential expression analysis using DESeq2 (Love et al., 2014). The
actual score was a sum of log₂ mean gene expression and value of DE test statistic minus a
threshold value that was a parameter to reflect importance of both basal expression level and
differential regulation of the enzyme. To assign a score to the nodes, we performed a t-test on
metabolite levels in two conditions, and used the logarithm of the p-value as the node score.
Metabolites that were not measured had a score of zero. In the scored network we used a heuristic
algorithm to find the most connected sub-network with maximal total node and edge scores.
Threshold values were selected to create a sub-network with 100-150 metabolites. This sub-
network was then enriched in reaction edges, connecting the metabolites provided the
corresponding enzymes were highly expressed (top 6000 by expression, irrespective of their
differential expression). Finally multiple atoms of the same compound were collapsed to yield a
network containing a single node for each compound to ease visualization and interpretation.
NSCLC patients and tissue samples
The study was approved by the local research ethics committee: UK National South West 4
Research Ethics Committee (REC) Southmead Hospital, Bristol BS10 5NB. REC no:
07/Q2002/6, South West 4 REC.
Three distinct samples of the tumour and adjacent normal tissue from the resection
margin were taken, flash frozen in liquid nitrogen, and stored at −80°C until further analysis.
Only those NSCLC tumour samples comprising at least 90% tumour tissue were analyzed.
Frozen tissue samples were homogenized and extracted as previously described. Briefly, a
Polytron homogenizer was used to generate tissue lysates in 1% NP40 lysis buffer. Protein
concentration was determined by BCA assay (Thermo Scientific, IL, USA) and tissue lysates
were stored at −80°C.
Supplementary References: Huang, S., Holzel, M., Knijnenburg, T., Schlicker, A., Roepman, P., McDermott, U., Garnett, M.,
Grernrum, W., Sun, C., Prahallad, A., et al. (2012). MED12 controls the response to multiple
cancer drugs through regulation of TGF-beta receptor signaling. Cell 151, 937-950.
Jha, A.K., Huang, S.C.-C., Sergushichev, A., Lampropoulou, V., Ivanova, Y., Loginicheva, E.,
Chmielewski, K., Stewart, K.M., Ashall, J., Everts, B., et al. (2015). Network integration of
parallel metabolomic-transcriptional data reveals novel metabolic modules regulating divergent
macrophage polarization. Immunity 42, 419-430.
Kanehisa, M., Goto, S., Sato, Y., Furumichi, M., and Tanabe, M. (2012). KEGG for integration
and interpretation of large-scale molecular data sets. Nucleic acids research 40, D109-114.
Love, M.I., Huber, W., and Anders, S. (2014). Moderated estimation of fold change and
dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550.